A conventional technique is known that simulates a face with makeup on a computer without actually applying makeup in order to sell commercial products for applying makeup. (See, for example, Patent Document 1.) However, according to Patent Document 1, a simulation result is displayed in a still image, which has made it difficult to check the made-up face reflecting a change in the user's expression. Therefore, the development of a technology that simulates makeup in a video having a change in a user's expression captured thereon has been advanced. (See, for Example, Patent Document 2.)
However, the makeup simulator described in Patent Document 2 calculates a makeup region to which to apply a face makeup by specifying a change in the user's expression by pixel regions corresponding to the mouth and eyes and tracking the pixel regions with template matching. (See, for example, paragraph [0028].) Tracking a change in the user's expression with the pixel regions corresponding to the mouth and eyes as described above imposes a large processing load on a computer, and thus has the problem of difficulty in accurately responding to cases such as eye closing.
[Patent Document 1] Japanese Laid-Open Patent Application No. 2001-346627
[Patent Document 2] Japanese Laid-Open Patent Application No. 2003-44837